Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

Recursion

(56,582 posts)
Wed Dec 18, 2013, 10:17 PM Dec 2013

Scientists Extract RSA Key From GnuPG Using Sound of CPU

Well, that's disconcerting...

http://it.slashdot.org/story/13/12/18/2122226/scientists-extract-rsa-key-from-gnupg-using-sound-of-cpu

In their research paper titled RSA Key Extraction via Low-Bandwidth Acoustic Cryptanalysis, Daniel Genkin, Adi Shamir and Eran Tromer et al. present a method for extracting decryption keys from the GnuPG security suite using an interesting side-channel attack. By analysing the acoustic sound made by the CPU they were able to extract a 4096-bit RSA key in about an hour (PDF). A modern mobile phone placed next to the computer is sufficient to carry out the attack, but up to four meters have been successfully tested using specially designed microphones.


12 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies
Scientists Extract RSA Key From GnuPG Using Sound of CPU (Original Post) Recursion Dec 2013 OP
Mindblowing. MannyGoldstein Dec 2013 #1
We Are All Unsafe From Prying Spies And Eyes cantbeserious Dec 2013 #2
I suspect this could be defeated. lapfog_1 Dec 2013 #3
The problem is that there is no actual random as far as computing concerned. Chan790 Dec 2013 #5
you misunderstood me lapfog_1 Dec 2013 #8
Kirchoff's postulate Recursion Dec 2013 #10
Just use a solid-state drive. nt jazzimov Dec 2013 #6
I read the story as they are analyzing the noise that the CPU makes as it calculates the encryption lapfog_1 Dec 2013 #9
Thats absolutely crazy NoOneMan Dec 2013 #4
Incidentally, that's basically how Gort resuscitated Klaatu after his fatal gunshot wound Orrex Dec 2013 #7
You think that one is weird? AgingAmerican Dec 2013 #11
Yeah, that was crazy Recursion Dec 2013 #12

lapfog_1

(29,205 posts)
3. I suspect this could be defeated.
Thu Dec 19, 2013, 12:11 AM
Dec 2013

What I would do is simply have the CPU perform random math calculation while doing the encryption routine (inject meaningless CPU cycles into the calculation, thus "noise" into the analysis).

while not perfect, it might make it very difficult to break via this method. However it would increase the cost (either by slowing down or using more CPU cycles) the encryption by using RSA (and likely most other public key encryption algorithms).

 

Chan790

(20,176 posts)
5. The problem is that there is no actual random as far as computing concerned.
Thu Dec 19, 2013, 12:28 AM
Dec 2013

Even random number generators are not truly random. They eventually fall into complex and predictable patterns once one discerns the formula used to generate the only-seemingly-random data unless the generation of numbers is based on an external random stimuli such as radioactive decay or micro-seismic activity rather than processor computations.

http://computer.howstuffworks.com/question697.htm

lapfog_1

(29,205 posts)
8. you misunderstood me
Thu Dec 19, 2013, 01:37 AM
Dec 2013

I would have the encryption routine, while doing actual encryption operations, select other operations at random (FFTs, etc) and run both operations at the same time (interleave the instruction streams). Just in case they can tell what part of L1 or L2 cache might be tied to what, I would flush cache at random intervals as well... and, just to confuse things further, I would randomly move chunks of memory around (in physical memory), just in case they can snoop that as well.

Recursion

(56,582 posts)
10. Kirchoff's postulate
Thu Dec 19, 2013, 03:41 AM
Dec 2013

Always assume an attacker has access to your entire cryptosystem except for the key. If you use a software random number generator, that's included.

lapfog_1

(29,205 posts)
9. I read the story as they are analyzing the noise that the CPU makes as it calculates the encryption
Thu Dec 19, 2013, 01:41 AM
Dec 2013

each math instruction, and the registers and circuits that do the math, might give off a signature noise, along with heat (and possibly a very small amount of light, along with the result of the calculation.

Latest Discussions»General Discussion»Scientists Extract RSA Ke...